feat(skills): add eval dataset for nemo-setup skill - #108
Open
tylersbray wants to merge 2 commits into
Open
Conversation
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
📝 WalkthroughWalkthroughAdds evaluation JSON fixture file for the Changesnemo-setup Skill Evaluation Dataset
Suggested reviewers
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
Contributor
|
Ship evals/evals.json for the NVSkills signoff process. The dataset contains 15 test cases (10 positive triggers, 5 negative) covering: - Fresh platform setup, bootstrap, and nemo setup flows - macOS gotchas (IPv6, unlinked-inode) - Port conflicts and database wipe procedures - Non-interactive CI mode - Switchyard middleware startup - Custom data directory configuration - InvalidTag encryption key desync recovery Negative cases confirm the skill stays silent on inference, customizer, evaluator, and other-product prompts. NVCARPS generates BENCHMARK.md from this dataset automatically. Ref: AIRCORE-692 Signed-off-by: Tyler Bray <tbray@nvidia.com>
tylersbray
force-pushed
the
AIRCORE-692-nemo-setup-evals/tbray
branch
from
May 29, 2026 18:53
5fec179 to
282b5ed
Compare
Contributor
|
/nvskills-ci |
mckornfield
approved these changes
Jun 22, 2026
| "question": "I want to run nemo setup on my Mac. Any gotchas I should know about?", | ||
| "expected_skill": "nemo-setup", | ||
| "expected_script": null, | ||
| "ground_truth": "On macOS, use http://127.0.0.1:8080 instead of localhost due to IPv6 resolution issues. Also be aware of the unlinked-inode gotcha: running `rm -rf ~/.local/share/nemo` while a nemo services run process is still alive does not reset state because the running process keeps writing to its open inode. Always kill the platform process first, then wipe the DB.", |
Contributor
There was a problem hiding this comment.
lol gotcha, maybe unlinked-inode issue ?
| "expected_script": null, | ||
| "ground_truth": "Switchyard is installed automatically by `make bootstrap-python` via the root workspace's enabled-plugins group. The library is vendored in-tree at plugins/nemo-switchyard/vendor/switchyard/. Start with debug logging: `LOG_LEVEL=DEBUG uv run nemo services run --services entities,models,inference-gateway,secrets --controllers models`. Switchyard is auto-discovered via its nemo.inference_middleware entry point.", | ||
| "expected_behavior": [ | ||
| "Note that Switchyard installs automatically via make bootstrap", |
Contributor
There was a problem hiding this comment.
should commands like this have `` or '' around them in the expected behavior sections?
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
skills/nemo-setup/evals/evals.json— the eval dataset required for NVSkills signoffContext
The NVSkills publishing pipeline requires an
evals/evals.jsonalongside each skill for the signoff process. NV-BASE uses this to evaluate trigger accuracy (discoverability), instruction-following (correctness), and effectiveness (with-skill vs without-skill).This dataset ships alongside
skills/nemo-setup/SKILL.mdwhich is being added in #104. The two PRs are additive and merge-order independent.Test plan
python -c "import json; json.load(open('skills/nemo-setup/evals/evals.json'))")/nvskills-ciafter feat: add repo-root skills for NVSkills catalog #104 mergesRef: AIRCORE-692
Summary by CodeRabbit